WO2016034516A2

WO2016034516A2 - Location and mapping device and method

Info

Publication number: WO2016034516A2
Application number: PCT/EP2015/069787
Authority: WO
Inventors: Guillaume TREHARD; Evangeline POLLARD; Fawzi Nashashibi; Benazouz Bradai
Original assignee: Valeo Schalter Und Sensoren Gmbh
Priority date: 2014-09-01
Filing date: 2015-08-28
Publication date: 2016-03-10
Also published as: FR3025325B1; US10379542B2; JP2017526083A; JP6682508B2; US20170261996A1; WO2016034516A3; FR3025325A1; CN107209515A; EP3189389A2; EP3189389B1

Abstract

The invention relates to a location and mapping device intended for being installed in a mobile vehicle in an environment comprising a plurality of zones, with: a receiving module (11) designed to receive data of the position (d(a)) of obstacles relative to the vehicle; a processing module (12, 14) designed to determine, in accordance with the data received (d(a)) and for a plurality of zones positioned relative to the vehicle, measurement values (M_SCANt,i) that respectively represent a likelihood of occupation of the zone in question, a likelihood of non-occupation of the zone in question and a likelihood of not knowing the zone in question; a mapping module (18, 20) designed to build, for each zone of said plurality of zones of the environment, values (M_GRIt,k) which respectively represent a likelihood of occupation of the zone in question, a likelihood of non-occupation of the zone in question and a likelihood of not knowing the zone in question; and a location module (16) designed to determine the position of the vehicle in the environment, which maximises a measurement of correspondence between the measurement values (M_SCANt,i) and the built values (M_GRIt,k). The invention also describes an associated method.

Description

Device and method for locating and mapping

TECHNICAL FIELD TO WHICH THE INVENTION RELATES The present invention relates to locating and mapping techniques used by a mobile machine in an environment.

It relates more particularly to a device and a method of locating and mapping for equipping such a mobile machine.

The invention is particularly advantageous in the case where a sensor delivers obstacle positioning data relative to the mobile machine.

BACKGROUND

Locating and mapping devices are known designed to equip a mobile machine (for example a robot) in order both to locate the mobile machine in its environment and to produce a cartography of this environment.

Such devices are commonly referred to as SLAM, with reference to the English acronym for "Simultaneous Localization and Mapping".

They are usually based on a probabilistic approach. In some solutions for example, it is determined, based on data provided by a sensor, the probability of occupation of each cell of a grid that covers the environment. One can refer to this topic in the article "Online localization and mapping with moving objects detection in dynamic outdoor environments", in IEEE 5th International Conference on Intelligent Computer Communication and Processing, 2009, ICCP 2009, 2009, pp. 410-408.

OBJECT OF THE INVENTION

In this context, the present invention proposes a locating and mapping device for equipping a mobile machine in an environment comprising a plurality of zones, with a receiving module designed to receive obstacle positioning data relative to the machine. ; a processing module designed to determine, based on the data received and for a plurality of zones positioned relative to the machine, measurement values respectively representative of a likelihood of occupation of the area concerned, a likelihood of non-occupation of concerned area and a likelihood of non-knowledge of the area concerned; a mapping module designed to construct, for each zone of said plurality of zones of the environment, values respectively representative of a likelihood of occupation of the zone concerned, of a likelihood of non-occupation of the zone concerned and a likelihood of non-knowledge of the area concerned; and a location module adapted to determine the positioning of the craft in the environment that maximizes a measure of correspondence between the measured values and the constructed values.

Thus, the position of the mobile machine is determined by finding the best match between the measurement values, generated on the basis of the positioning data received from a sensor at a given time, and the constructed values, which describe the mapping. developed by the device at previous iterations.

These values, however, are not limited to evaluating, like classical probabilistic approaches, the probability of presence or absence of an obstacle in the different zones of the environment, but also take into account the lack of information for certain areas, thanks to the representative values of the likelihood of non-knowledge of these zones.

It should be noted that the values representative of the likelihood of occupation of a zone and the likelihood of non-occupation of a zone are also influenced by the absence of information for the area concerned, since in general, the above-mentioned three representative values (likelihood of occupation, likelihood of non-occupation and likelihood of non-knowledge), referred to as "masses" in the description that follows, have a fixed sum (generally equal to 1).

According to other optional features (and therefore not limiting):

the location module is designed to determine the correspondence measurement by using a measure of conflict between the measurement values and the constructed values;

the correspondence measurement is indicative of a correspondence between a spatial distribution of the measurement values representative of the likelihood of occupation and a spatial distribution of the constructed values representative of the likelihood of occupation; the locating module is designed to determine the measurement of correspondence for a given positioning of the mobile machine by using a sum of terms each associated with a given zone of the environment, and to calculate each term using the product of the measurement value representative of the likelihood of occupation of the zone corresponding to the given zone for the given positioning of the machine and the constructed value representative of the likelihood of occupation of the given zone;

the location module is designed to calculate each term by dividing the aforementioned product by a conflict factor using a sum of the measurement value representative of the likelihood of occupation of the zone corresponding to the given zone for the given positioning of the machine, multiplied by the constructed value representative of the likelihood of non-occupation of the given zone, and the measurement value representative of the likelihood of non-occupation of the zone corresponding to the given zone for the given positioning of the gear, multiplied by the constructed value representative of the likelihood of occupation of the given area;

the mapping module comprises a combination module designed to determine for each zone, by combining the said measurement values relative to the zone concerned and the values constructed by the mapping module for the zone concerned during a previous iteration, values raw respectively representative of a likelihood of occupation of the area concerned, a likelihood of non-occupation of the area concerned, a likelihood of non-knowledge of the area concerned and a likelihood of conflict in the area concerned;

the mapping module comprises a normalization module designed to determine the said constructed values according to the said raw values;

the processing module comprises a conversion module designed to determine, for each zone of the plurality of zones positioned with respect to the machine, said measurement values relating to the zone concerned as a function of corresponding values assigned, in a polar grid; , to a cell corresponding to the area concerned.

The invention also proposes a system comprising a locating and mapping device as mentioned above and a sensor designed to generate said positioning data. This sensor is for example a lidar, but it could be another type of sensor, for example a radar.

The invention proposes, in a correlative manner, a method of locating and mapping used by a mobile machine in an environment comprising a plurality of zones, comprising the following steps:

receiving obstacle positioning data relating to the machine;

determination, based on the data received and for a plurality of zones positioned with respect to the machine, of measurement values respectively representative of a likelihood of occupation of the zone concerned, of a likelihood of non-occupation of the concerned area and a likelihood of non-knowledge of the area concerned;

constructing, for each zone of said plurality of zones of the environment, respectively representative values of a likelihood of occupation of the zone concerned, a likelihood of non-occupation of the zone concerned and a likelihood of lack of knowledge of the area concerned;

- Locating the machine by determining the positioning of the machine in the environment that maximizes a measure of correspondence between the measured values and the constructed values.

This method may optionally further comprise a step of combining said measurement values relative to the area concerned and of values constructed by the mapping module for the area concerned during a previous iteration in order to generate raw values respectively representative of a given region. likelihood of occupation of the area concerned, a likelihood of non-occupation of the area concerned, a likelihood of non-knowledge of the area concerned and a likelihood of conflict in the area concerned.

The optional features presented above relating to the locating and mapping device can also be applied to such methods.

DETAILED DESCRIPTION OF AN EXEMPLARY EMBODIMENT The following description with reference to the accompanying drawings, given by way of non-limiting examples, will make it clear what the invention consists of and how it can be implemented. In the accompanying drawings:

FIG. 1 represents a system comprising a sensor and an exemplary localization and mapping device according to the invention;

FIG. 2 represents in functional form the processes performed by the localization and mapping device of FIG. 1;

FIG. 3 represents a polar grid constructed from the positioning data received from the sensor.

FIG. 1 schematically represents the main elements of a system including a locating and mapping device 10.

Such a system equips a mobile machine in an environment, for example a motor vehicle or a robot. As will be apparent from the description which follows, a system equipped with a localization and mapping device, sometimes called SLAM (Anglo-Saxon acronym meaning "Simultaneous Location And Mapping"), allows both the location of the system (and therefore mobile gear) in the environment and the discovery of environmental mapping.

The system of FIG. 1 comprises a sensor 2 which generates positioning data (relative to the system) of obstacles present in the environment. The sensor 2 is for example lidar type; in this case, the positioning data are representative, for a plurality of orientations with respect to a reference direction of the system, of the distance d (a) (with respect to the system) of the first obstacle encountered in the direction associated with this orientation a.

One could alternatively use a sensor of another type, for example a radar.

The system of FIG. 1 also comprises the locating and mapping device 10, which receives as input the positioning data generated by the sensor 2 and generates at the output location information X _t of the system (and therefore of the mobile machine equipped with the system) at each moment t and GRI mapping information.

This GRI mapping information is for example an occupancy grid which describes, in the form of a grid, the locations of the environment in which an obstacle has been detected.

In the case described here where the environment is mapped in 2 dimensions, the location information X _t include for example the x, y coordinates of the system in the mapping plane (precisely its x-coordinate and its y-coordinate) and its heading Θ (in English "heading"). This location information X is determined as described below with respect to initial coordinates x ₀ , y _o and an initial heading θ ₀ defined by convention (for example, x ₀ = 0; y ₀ = 0; θ ₀ = 0).

In order to implement the processes described below, the locating and mapping device 10 here comprises a processor 4 (for example a microprocessor) and a storage means 6 (for example a hard disk).

The processor 4 is designed to implement the processes described below, in particular with reference to FIG. 2, for example because of the execution of computer program instructions stored in the storage means 6. In a variant these processing could be carried out by a specific application integrated circuit (or ASIC of the "Application Specifies Integrated Circuit") implanted in the locating and mapping device 10.

The storage means 6 also allow the storage of the data used during the processing described below, in particular the data representative of likelihoods or credibilities, hereinafter called "masses".

FIG. 2 represents in functional form the processes carried out by the locating and mapping device 10 in order to determine the location information X _t and the mapping information GRI from the positioning data d (a) generated by the sensor 2.

Thus, the various treatments performed by the locating and mapping device 10 are shown in FIG. 2 implemented by different modules 1 1, 12, 14, 1 6, 18, 20, 22 although these can be set implemented in practice because of the execution of instructions by the same processor 4.

The locating and mapping device 10 comprises a reception module 1 1 connected to the sensor 2 and which thus receives the positioning data d (a) generated by the sensor 2 and transmits them to a polar grid construction module 12.

According to the positioning data d (a) that it receives as input at a time t, the polar grid construction module 12 determines, for each of the cells j of a polar grid, the three following parameters:

an occupancy mass M_POL _{t, j} (0) (in English "occupation mass");

a mass of non-occupation M_POL _{t, j} (F) (in English "free mass");

a mass of non-knowledge M_POL _{t, j} (U) (in English "unknown mass").

For each cell j, the sum of the different masses relating to this cell is equal to 1.

As shown in FIG. 3, a polar grid is a set of cells of the environment (here represented in two dimensions) in which each cell corresponds to a given range of distances to the system (specifically to sensor 2) and to a given angular sector. (the angles being measured with respect to the reference direction of the system already mentioned).

The occupancy mass M_POL _{t, j} (0) represents the possibility that the cell j concerned in the polar grid is occupied, according to some information, here the positioning data received from the sensor 2 at time t. The occupation mass M_POL _t , j (0) is therefore representative of a likelihood or credibility of occupation of this cell j.

The non-occupation mass M_POL _{t, j} (F) represents the possibility that the cell j concerned in the polar grid is empty, according to some information, here the positioning data received from the sensor 2 at time t. The non-occupation mass M_POL _t , j (F) is therefore representative of a likelihood or credibility of non-occupation of this cell j.

The mass of non-knowledge M_POL _{t, j} (U) represents the possibility that the cell j concerned has not been explored, that is to say that no knowledge is available for this cell according to the information available (here the positioning data). The mass of non-knowledge M_POL _{t, j} (U) is therefore representative of a likelihood or credibility of non-knowledge of the cell j.

For cells j situated in the angular sector corresponding to a direction Θ, the following are for example determined as follows:

for the cell j (blackened in FIG. 3) located at a distance d (6) from the system (that is to say, in the case where the sensor 2 is a lidar, the one where an object has been impacted by the laser shown in dashed lines in FIG.

M_POL _{t, j} (0) = I CONF, M_POL _{t, j} (F) = 0, M_POL _{t, j} (U) = 1 - ICONF, WHERE I _CONF is a confidence index of the sensor 2 (for example predefined or delivered by the sensor 2), typically between 0.5 and 1;

for cells j situated at a distance less than the distance d (6) (that is to say those traversed by the laser in the case of lidar),

for cells j located at a distance greater than the distance d (6) (that is to say those situated beyond the impact of the laser in the case of the lidar, represented hatched in FIG. 3),

When the sensor delivers several positioning data d (6) for the same angular sector of the polar grid (for example when the lidar delivers data for several altitudes or layers, or when the resolution of the lidar is thinner than that of the grid polar), for example, the smallest distance d (6) for the determination of masses as described above is retained.

The polar grid construction module 12 transmits all the mass parameters M_POL _{t, j} thus determined to a conversion module 14 which assigns, for each cell j of the polar gate, the masses determined for this cell j to one ( or more) corresponding cell (s) i of a measurement grid in Cartesian coordinates (expressed relative to the current position of the system at time t).

In other words, for everything

when the cell i of the measurement grid corresponds to the cell j of the polar grid.

We thus define for each cell i of the measurement grid:

an occupancy mass M_SCAN _{t, j} (O) for the cell i, representing the likelihood (or credibility) of occupation of the cell i (according to the positioning data provided by the sensor 2 at time t );

a non-occupation mass M_SCAN _{t, j} (F) for the cell i, representing the likelihood of non-occupation of the cell i (according to the positioning data provided by the sensor 2 at time t);

a mass of non-knowledge M_SCAN _{t, j} (U) for cell i, representing the likelihood of non-knowledge of cell i (according to the positioning data provided by sensor 2 at time t).

For each cell i, the sum of the different masses relating to this cell is equal to 1.

Here, the term "processing module" is used to denote the assembly formed by the polar grid construction module 12 and the conversion module 14. The processing module thus makes it possible to determine the masses M_SCAN _{t, j} (A) from the data of FIG. positioning d (θ).

As can be seen in FIG. 2, the different masses M_SCAN _{t, j} (A) (with A € {O, F, U}) produced by the conversion module 1 6 for the different cells i of the measurement grid are transmitted from a part to a location module 1 6 and secondly to a combination module 18.

The location module 1 6 also receives as input the following information produced by the combination module 18 and the normalization module 20 at the previous iteration t-1 (as explained hereinafter for the current iteration t) and relative to the cells k of a grid representing the environment (in Cartesian coordinates with respect to a fixed reference, such as the initial position of the system):

for each cell k, an occupation mass M_GRI _t-1 , k (0), representing the likelihood of occupation of the cell k, constructed by the device 10 due to the processing carried out up to the instant t- 1;

for each cell k, a non-occupation mass M_GRI _t-1 , k (F), representing the likelihood of non-occupation of the cell k, constructed by the device 10 due to the processing carried out up to the moment t-1;

for each cell k, a mass of non-knowledge M_GRI _t-1 , k (U), representing the likelihood of non-knowledge of the cell k, constructed by the device 10 due to the processing carried out up to the moment t-1.

Note that this information is not available during the first iteration. However, no processing is necessary during this first iteration: the location module 1 6 then only indicates that the current position (defined by the coordinates x, y and heading Θ) is equal to the initial position already mentioned (defined by the coordinates x ₀ , y _o and by the heading θο).

For the other iterations, the location module 1 6 determines the position X _t of the system (and therefore of the mobile machine), defined here by the coordinates x, y and the heading Θ, which maximizes a measurement of correspondence between the values associated with the measurement grid (located in relation to the position of the system) and the values associated with the fixed grid representing the environment.

For example, the following measurement is used for this purpose, defined for a Z position by:

where i (k, Z) is the cell of the measurement grid (positioned relative to the system, that is to say to the mobile machine as already indicated) corresponding to the cell k of the fixed gate when the system has the Z position and where M _{t, k, z} (C) measures the conflict between the information used in the numerator:

A correspondence measurement is thus used which quantifies a correspondence between a spatial distribution (within the cells) of the occupation likelihoods in the measurement grid and a spatial distribution of the occupation likelihoods in the fixed grid.

The use of the denominator (1 - M _{t, k, z} (C)) makes it possible to weight in the sum of the cells for which the information used in the numerator is contradictory in order to take it into account wisely.

As a variant, it would be possible to use, for each term of the sum mentioned above, another fusion operator in order to combine the masses M_GRI _{t-1, k} received from the combination module 18 (via the normalization module 20) and the M_SCAN masses _{t, j (k, z)} received from the conversion module 14, for example a fusion operator such as those used in the context of the theory of evidence (also called Dempster-Shafer theory).

The value X _T delivered by the location module 16 is therefore in this case the value of Z which maximizes the measurement N (Z): X is such that N (X) = maxz N (Z).

This is for example determined in practice by calculating the measurement N (Z) for a plurality of Z positions. In the following, z _x , z _y , ζ _{θ are} respectively denoted by the abscissa, the ordinate and the heading defining a Z position. .

According to a first embodiment, the measurement N (Z) is calculated for a (discrete) set of Z positions verifying:

where x _t-1 , y _t-1 , Θ _t-1 are the abscissa, the ordinate and the heading defining the position X _t-1 determined at the preceding iteration (or the abscissa, the ordinate and the initial captions during the iteration t = 1) and where S _x , S _y , S _θ are predetermined thresholds. By expressing the coordinates in number of steps of the fixed grid (the cells being here of the same size in the fixed grid and in the measuring grid) - or in degrees for the heading, we use for example S _x = S _y = 5 and S _θ = 10 ° and:

a calculation of the measure N (Z) for 1331 possible positions Z.

According to a second embodiment, a priori position is used

derived from the position

determined at the previous iteration on the basis of a constant speed model:

where is the derivative of the position X _t-1 at time t-1

We can then calculate the measure N (Z) for a (discrete) set of Z positions verifying:

where S _x , S _y , S _θ are predetermined thresholds.

By expressing the coordinates in number of steps of the fixed gate (the cells being here of the same size in the fixed gate and in the measuring gate), for example S _x = S _y = 5 and S _θ = 10 ° are used. :

a calculation of the measurement N (Z) for 1331 possible positions Z around the position a priori determined on the basis of a speed model

constant.

Reference can be made to this topic in the article "A real-time robust SLAM for large-scale outdoor environments" by J. Xie, F. Nashashibi, MN Parent and O. Garcia-Favrot in ITS World Congress. 2010. The position X _t determined by the locator module 1 6 is transmitted to the combination module 18, which also receives as input as already indicated the different masses produced by the

conversion module 14 for the different cells i of the measurement grid, as well as the different masses produced as

explained above by the normalization module 20 at the previous iteration t-1 for the cells k of the fixed gate.

The combination module 18 then combines the information provided by the sensor 2 at the instant t (represented by the masses M_SCAN _{t, i} (A)) and the information constructed by the device 10 up to the instant t-1 ( represented by the masses M_GRI _{t-1, k} (A)) in order to derive from them raw masses M_COMB _{t, k} (A) representative of the state of knowledge of the fixed gate at time t, for example in applying to each cell k of the fixed grid a formula of the type:

where i (k, X _t ) is the cell of the measurement grid which corresponds to the cell k of the fixed gate when the system is at the position X _t and where is a

Conjunction defined for example in the article "Credibilist occupancy grids for vehicle perception in dynamic environments", J. Moras, V. Cherfaoui and P. Bonnifait in IEEE Int. Conf. Robot. Automat. (ICRA201 1), pages 84-89, 201 1.

By explaining the calculations made because of this operator we obtain

so for each cell k:

a raw mass of occupation M_COMB _{t, k} (O), representing the likelihood of occupation of the cell k, here by:

a non-occupation gross mass M_COMB _{t, k} (F), representing the likelihood of non-occupation of the cell k, here by:

a gross mass of non-knowledge M_COMB _{t, k} (U), representing the likelihood of non-knowledge of the cell k, here by:

a raw mass of conflict M_COMB _{t, k} (C), representing a likelihood of conflict about this cell k (between the information delivered by the sensor 2 and the information constructed by the device 10 until the instant t-1), here by:

For each cell k, the sum of the different raw masses relating to this cell is equal to 1.

The raw masses M_COMB _{t, k} produced by the combination module 18 are transmitted to the normalization module 20 which retains (in each cell) the distribution of the masses of occupation, non-occupation and non-knowledge but eliminates the notion of mass of conflict, which makes it possible to free oneself from moving obstacles. According to one conceivable embodiment, the conflict raw masses M_COMB _{t, k} (C) for the different cells k can be stored in order to be used for the detection of moving obstacles.

The normalization module 20 thus calculates the following masses (after normalization):

for all cells k.

We thus obtain for each cell k a mass of occupation

M_GRI _{t, k} (0), a non-occupancy mass M_GRI _{t, k} (F) and a mass of non-knowledge M_GRI _{t, k} (U), which here constitute the mapping information GRI delivered by the device 10. Thanks to the normalization, for each cell k, the sum of the different masses

relating to this cell is equal to 1.

Here, the term "mapping module" refers to the assembly formed by the combination module 18 and the normalization module 20.

We can then for example represent this information on a map (displayed for example on a screen of the mobile machine), possibly by showing each cell k in a color defined according to the masses.

relating to this cell k; for example, for a cell k, a color is used whose trichromatic coordinates in the reference frame RGB are

As already indicated, the masses M_GRI _{t, k} produced at the output of the module 20 are transmitted (through a delay of an iteration 22) to the location module 1 6 and the combination module 18 to be used at the next iteration.

Claims

1. Locating and mapping device (10) for equipping a mobile machine in an environment comprising a plurality of zones, with:

- a receiving module (1 1) designed to receive positioning data (d (a)) obstacles relative to the machine;

- a processing module (12, 14) designed to determine, based on the data received (d (a)) and for a plurality of zones positioned relative to the machine, measurement values (M_SCAN _{t, j} ) respectively representative of a likelihood of occupation of the area concerned, a likelihood of non-occupation of the area concerned and a likelihood of non-knowledge of the area concerned;

a mapping module (18, 20) designed to construct, for each zone of said plurality of zones of the environment, values (M_GRI _t , _k ) respectively representative of a likelihood of occupation of the zone concerned, d a likelihood of non-occupation of the area concerned and a likelihood of non-knowledge of the area concerned; and

a locating module (1 6) designed to determine the positioning of the machine in the environment which maximizes a measurement of correspondence between the measured values (M_SCAN _{t, j} ) and the constructed values (M_GRI _t , k) -

A locator and mapping device according to claim 1, wherein the locator module (1 6) is adapted to determine the mapping measure using a measure of conflict between the measured values (M_SCAN _{t, j} ) and the constructed values (M_GRI _{t, k} ).

A locating and mapping device according to claim 1 or 2, wherein the correspondence measure is indicative of a correspondence between a spatial distribution of the measured values (M_SCAN _{t, j} ) representative of the likelihood of occupancy and a spatial distribution of the constructed values (M_GRI _t , _k ) representative of the likelihood of occupation.

4. Locating and mapping device according to one of claims 1 to 3, wherein the location module (1 6) is designed to determining the correspondence measurement for a given positioning of the mobile machine by using a sum of terms each associated with a given area of the environment, and for calculating each term using the product of the measurement value (M_SCAN _{t, i} ) representative of the likelihood of occupation of the zone corresponding to the given zone for the given positioning of the machine and the constructed value (M_GRI _t , _k ) representative of the likelihood of occupation of the given zone.

A locating and mapping device according to claim 4, wherein the locator module (16) is adapted to calculate each term by dividing the aforementioned product by a conflict factor using a sum of the measurement value (M_SCAN _{t, j} ) representative of the likelihood of occupation of the zone corresponding to the given zone for the given positioning of the machine, multiplied by the constructed value (M_GRI _t , _k ) representative of the likelihood of non-occupation of the given zone, and the measurement value (M_SCAN _{t, j} ) representative of the likelihood of non-occupation of the zone corresponding to the given zone for the given positioning of the machine, multiplied by the constructed value (M_GRI _t , _k ) representative of the likelihood of occupation of the given area.

6. A locating and mapping device according to one of claims 1 to 5, wherein the mapping module (18, 20) comprises a combination module (18) designed to determine for each zone, combining said measured values. (M_SCAN _t , i) relative to the area concerned and constructed values (M_GRI _t-1 , _k ) by the mapping module (18, 20) for the zone concerned during a previous iteration, raw values (M_COMB _t , _k ) respectively representative of a likelihood of occupation of the area concerned, a likelihood of non-occupation of the area concerned, a likelihood of non-knowledge of the area concerned and a likelihood of conflict on the area concerned.

The locating and mapping device according to claim 6, wherein the mapping module (18, 20) comprises a normalization module (20) adapted to determine said constructed values (M_GRI _{t, k} ). according to said raw values (M_COMB _{t, k} ) -

8. Locating and mapping device according to one of claims 1 to 7, wherein the processing module (12, 14) comprises a conversion module (14) adapted to determine, for each zone of the plurality of positioned zones relative to the machine, said measurement values (M_SCAN _{t, j} ) relative to the area concerned as a function of corresponding values (M_POL _{t, j} ) allocated, in a polar grid, to a cell corresponding to the zone concerned.

9. A system comprising a locating and mapping device (10) according to one of claims 1 to 8 and a sensor (2) adapted to generate said positioning data (d (a)).

The system of claim 9, wherein the sensor (2) is a lidar.

1 1. A method of locating and mapping used by a mobile craft in an environment comprising a plurality of areas, comprising the steps of:

receiving positioning data (d (a)) of obstacles relative to the machine;

determination, as a function of the data received (d (a)) and for a plurality of zones positioned with respect to the machine, of measurement values (M_SCAN _{t, j} ) respectively representative of a likelihood of occupation of the zone concerned, a likelihood of non-occupation of the area concerned and a likelihood of non-knowledge of the area concerned;

- construction, for each zone of said plurality of zones of the environment, values (M_GRI _{t, k} ) respectively representative of a likelihood of occupation of the area concerned, a likelihood of non-occupation of the area concerned and a likelihood of non-knowledge of the area concerned;

locating the machine by determining the positioning of the machine in the environment which maximizes a measurement of correspondence between the measurement values (M_SCAN _{t, j} ) and constructed values (M_GRI _t , _k ) -

12. The method of claim 1 1, comprising a step of combining, for each zone of the environment, said measurement values (M_SCAN _{t, j} ) relative to the area concerned and values (M_GRI _{t - 1, k} ). constructed by the mapping module for the area concerned during a previous iteration in order to generate raw values (M_COMB _t , _k ) respectively representative of a likelihood of occupation of the zone concerned, of a likelihood of non-occupation the area concerned, a likelihood of non-knowledge of the area concerned and a likelihood of conflict in the area concerned.